tasks
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements strict guidelines for constructing shell commands using quoted heredocs and unique delimiters to prevent command injection when handling user-provided text.
- [SAFE]: The instructions mandate the use of Markdown validation and explicitly forbid raw HTML, which ensures data consistency and prevents potential cross-site scripting (XSS) or rendering vulnerabilities in downstream viewers.
- [SAFE]: The shell script
scripts/tasks.shis a benign wrapper that delegates execution to the local system binaryttasks. - [SAFE]: The skill includes behavioral constraints ('Observed-Failure Counters') that guide the agent to provide complete, verified information, effectively mitigating the risk of providing provisional or misleading data.
Audit Metadata